Statement on AI

Unless otherwise stated, everything you read here is written and created by me. Furthermore, I do not use AI-generated code on my sites. I write to think and the act of creation is the point, not the output. I have no interest in using generative AI to write because tools to automate the creative process are anathema to my goals for this site.

That said, I do sometimes work with open-source, auditable, language models and bespoke domain-specific language models made from public domain sources in my personal research:

1. As exploratory tools as a tool to generate metadata for and finding patterns or outliers in sets of data that are too big for me to review by myself. The quality of the analysis is always reviewed and verified. I will always disclose when and how their product has contributed to my work and research.
2. Build and maintain first-hand experience and knowledge using language models to better understand the state of the field and how they are being used in practice.
3. I also sometimes use machine learning tools to help transcribe audio recordings, mostly for my own reference needs. I personally review all of them for accuracy.
4. Regardless of use, I only use tools with clear privacy rules and verified data sovereignty rules where no data is retained and used to train third-party AI resources.

I have also written more in-depth on my stance on the use of LLMs and AI on my blog.